U.S. patent application number 15/795537 was filed with the patent office on 2018-03-08 for application transparent continuous availability using synchronous replication across data stores in a failover cluster.
The applicant listed for this patent is MICROSOFT TECHNOLOGY LICENSING, LLC. Invention is credited to Roopesh Battepati, Vyacheslav Kuznetsov, Ganesh Prasad.
Application Number | 20180067698 15/795537 |
Document ID | / |
Family ID | 54207728 |
Filed Date | 2018-03-08 |
United States Patent
Application |
20180067698 |
Kind Code |
A1 |
Prasad; Ganesh ; et
al. |
March 8, 2018 |
APPLICATION TRANSPARENT CONTINUOUS AVAILABILITY USING SYNCHRONOUS
REPLICATION ACROSS DATA STORES IN A FAILOVER CLUSTER
Abstract
Disclosed herein is a system and method for automatically moving
an application from one site to another site in the event of a
disaster. Prior to coming back online the application is configured
with information to allow it to run on the new site without having
to perform the configuration actions after the application has come
online. This enables a seamless experience to the user of the
application while also reducing the associated downtime for the
application.
Inventors: |
Prasad; Ganesh; (Bellevue,
WA) ; Battepati; Roopesh; (Bellevue, WA) ;
Kuznetsov; Vyacheslav; (Sammamish, WA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
MICROSOFT TECHNOLOGY LICENSING, LLC |
Redmond |
WA |
US |
|
|
Family ID: |
54207728 |
Appl. No.: |
15/795537 |
Filed: |
October 27, 2017 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14848318 |
Sep 8, 2015 |
9804802 |
|
|
15795537 |
|
|
|
|
62047634 |
Sep 8, 2014 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G06F 3/0683 20130101;
G06F 3/0619 20130101; G06F 11/2033 20130101; G06F 3/065 20130101;
G06F 11/2041 20130101; G06F 11/2048 20130101; G06F 11/2097
20130101; G06F 11/203 20130101 |
International
Class: |
G06F 3/06 20060101
G06F003/06; G06F 11/20 20060101 G06F011/20 |
Claims
1. A method for failing over an application from a failed primary
site to a replication site, the method comprising: determining that
the primary site has failed, based at least on one or more primary
nodes of the primary site not being connected to an application
cluster resource group (ACRG) at the primary site; based at least
on the determination that the primary site has failed: building a
candidate list of a plurality of replication cluster resource
groups (RCRG) at secondary sites that are each capable of becoming
a new ACRG; selecting a particular RCRG at a particular secondary
site from the candidate list, based at least on removing another
RCRG at another secondary site from the candidate list; converting
the particular RCRG to be the new ACRG; and designating the
particular secondary site as a new primary site; and prior to
bringing the application online, configuring the application to
execute on one or more new primary nodes and the new ACRG at the
new primary site, including configuring a mapping of a cluster
physical disk resource to a physical data store.
2. The method of claim 1, wherein removing the other RCRG at the
other secondary site from the candidate list is based at least on a
failure to establish a connection to the other RCRG.
3. The method of claim 1, wherein removing the other RCRG at the
other secondary site from the candidate list is based at least on
an analysis of a replication status of each RCRG.
4. The method of claim 3, wherein the replication status of each
RCRG indicates whether each RCRG is in sync status, or is not in
sync status.
5. The method of claim 1, wherein the particular RCRG at the
particular secondary site is selected from the candidate list based
upon gathered characteristics for each candidate in the candidate
list.
6. The method of claim 5, wherein the gathered characteristics
comprise one or more of geographical location, performance
characteristics, or quality characteristics.
7. The method of claim 1, wherein the particular RCRG at the
particular secondary site is selected from the candidate list based
upon a geographical constraint.
8. A computer system, comprising: one or more processors; and one
or more computer-readable media having stored thereon
computer-executable instructions that are executable by the one or
more processors to cause the computer system to fail over an
application from a failed primary site to a replication site, the
computer-executable instructions being executable to cause the
computer system to at least: determine that the primary site has
failed, based at least on one or more primary nodes of the primary
site not being connected to an application cluster resource group
(ACRG) at the primary site; based at least on the determination
that the primary site has failed: build a candidate list of a
plurality of replication cluster resource groups (RCRG) at
secondary sites that are each capable of becoming a new ACRG;
select a particular RCRG at a particular secondary site from the
candidate list, based at least on removing another RCRG at another
secondary site from the candidate list for failure to establish a
connection to the other RCRG; convert the particular RCRG to be the
new ACRG; and designate the particular secondary site as a new
primary site; and prior to bringing the application online,
configure the application to execute on one or more new primary
nodes and the new ACRG at the new primary site, including
configuring a mapping of a cluster physical disk resource to a
physical data store.
9. The computer system of claim 8, wherein removing the other RCRG
at the other secondary site from the candidate list is based at
least on an analysis of a replication status of each RCRG.
10. The computer system of claim 9, wherein the replication status
of each RCRG indicates whether each RCRG is in sync, or is not in
sync.
11. The computer system of claim 8, wherein the particular RCRG at
the particular secondary site is selected from the candidate list
based upon gathered characteristics for each candidate in the
candidate list.
12. The computer system of claim 11, wherein the gathered
characteristics comprise one or more of geographical location,
performance characteristics, or quality characteristics.
13. The computer system of claim 8, wherein the particular RCRG at
the particular secondary site is selected from the candidate list
based upon a geographical constraint.
14. The computer system of claim 8, wherein the particular RCRG at
the particular secondary site is selected from the candidate list
based upon a random selection.
15. A computer program product comprising one or more hardware
storage devices having stored thereon computer-executable
instructions that are executable by one or more processors to cause
a computer system to fail over an application from a failed primary
site to a replication site, the computer-executable instructions
being executable to cause the computer system to at least:
determine that the primary site has failed, based at least on one
or more primary nodes of the primary site not being connected to an
application cluster resource group (ACRG) at the primary site;
based at least on the determination that the primary site has
failed: build a candidate list of a plurality of replication
cluster resource groups (RCRG) at secondary sites that are each
capable of becoming a new ACRG; select a particular RCRG at a
particular secondary site from the candidate list, based at least
on removing another RCRG at another secondary site from the
candidate list for having a not in sync status; convert the
particular RCRG to be the new ACRG; and designate the particular
secondary site as a new primary site; and prior to bringing the
application online, configure the application to execute on one or
more new primary nodes and the new ACRG at the new primary site,
including configuring a mapping of a cluster physical disk resource
to a physical data store.
16. The computer program product of claim 15, wherein removing the
other RCRG at the other secondary site from the candidate list is
based at least on a failure to establish a connection to the other
RCRG.
17. The computer program product of claim 15, wherein a replication
status of each RCRG indicates whether each RCRG is in sync, or is
not in sync.
18. The computer program product of claim 15, wherein the
particular RCRG at the particular secondary site is selected from
the candidate list based upon gathered characteristics for each
candidate in the candidate list.
19. The computer program product of claim 18, wherein the gathered
characteristics comprise one or more of geographical location, a
performance characteristic, or a quality characteristic.
20. The computer program product of claim 15, wherein the
particular RCRG at the particular secondary site is selected from
the candidate list based upon a geographical constraint.
Description
CROSS REFERENCE TO RELATED APPLICATION
[0001] This application is a continuation of U.S. patent Ser. No.
14/848,318, filed Sep. 8, 2015, and titled "APPLICATION TRANSPARENT
CONTINUOUS AVAILABILITY USING SYNCHRONOUS REPLICATION ACROSS DATA
STORES IN A FAILOVER CLUSTER," and which is now U.S. Pat. No.
9,804,802, issued Oct. 31, 2017. The entire contents of the
foregoing application are incorporated by reference herein in their
entirety.
BACKGROUND
[0002] Applications and sites fail for a variety of reasons. When
they fail, it becomes necessary to move the application to a new
location to maintain application availability. Synchronous block
replication in a failover cluster environment requires application
downtime and manual storage resource dependency changes as a part
of a disaster recovery workflow. This is because the application is
moved from the location that has failed to another location that is
capable of supporting the application. In order to achieve this,
the physical disk resource to physical data store mapping needs to
be changed to permit the associated application to operate on the
new location. These changes are made after the associated
application has been brought back up on the new location. This
results in an extended period of application downtime for the
user.
SUMMARY
[0003] The following presents a simplified summary of the
disclosure in order to provide a basic understanding to the reader.
This summary is not an extensive overview of the disclosure and it
does not identify key/critical elements of the invention or
delineate the scope of the invention. Its sole purpose is to
present some concepts disclosed herein in a simplified form as a
prelude to the more detailed description that is presented
later.
[0004] The present disclosure provides a system and method for
automatically moving an application from one site to another site
in the event of a disaster. Prior to coming back online, the
application is configured with information to allow it to run on
the new site without having to perform those actions after the
application has become online. This provides for a seamless
experience to the user of the application while also reducing the
associated downtime for the application.
[0005] When a primary site for an application goes down due to
disaster or other reason and the application resource group (that
also contains any replicated disks) moves to a secondary site that
holds a synchronous target replica, the cluster physical disk
resource to physical data store mapping is modified to use the
target replica before the application resource comes online
resulting in automatic role switch. This mechanism virtualizes the
application dependent cluster physical disk resource from multiple
synchronous copies of data in various sites allowing seamless
failover and failback capabilities.
[0006] The present disclosure also allows a cluster replication
resource to maintain cluster wide replication state of all target
replicas allowing it to decide if a target is eligible to be source
of replication in the event of a disaster. The target replica
connects to source replica without using a well-known endpoint.
When source replica fails over to a different node within a primary
site, the target replica in secondary site discovers the new
endpoint to connect to and resumes replication.
[0007] The cluster replication resource automatically adjusts the
possible owners nodes of the source and target replica based on the
replication state, replication service availability, storage
connectivity, arrival, or departure of nodes in cluster due to
membership changes. This allows the application resource group to
failover to only those nodes where this is a high chance of the
success due to availability of all required resources.
[0008] Many of the attendant features will be more readily
appreciated as the same becomes better understood by reference to
the following detailed description considered in connection with
the accompanying drawings.
DESCRIPTION OF THE DRAWINGS
[0009] The present description will be better understood from the
following detailed description read in light of the accompanying
drawings, wherein:
[0010] FIG. 1 is a block diagram illustrating a system 100 for
providing application transparent continuous availability using
synchronous replication across data stores in a failover cluster
according to an illustrative embodiment.
[0011] FIG. 2 is a flow diagram illustrating a process for
selecting a secondary replication group and automatically
performing role switching according to one illustrative
embodiment.
[0012] FIG. 3 is a flow diagram illustrating a process for
switching roles according to one embodiment.
[0013] FIG. 4 is a block diagram illustrating a computing device
that can implement the enhanced indexing system according to one
embodiment.
[0014] Like reference numerals are used to designate like parts in
the accompanying drawings.
DETAILED DESCRIPTION
[0015] The detailed description provided below in connection with
the appended drawings is intended as a description of the present
examples and is not intended to represent the only forms in which
the present example may be constructed or utilized. The description
sets forth the functions of the example and the sequence of steps
for constructing and operating the example. However, the same or
equivalent functions and sequences may be accomplished by different
examples.
[0016] When elements are referred to as being "connected" or
"coupled," the elements can be directly connected or coupled
together, or one or more intervening elements may also be present.
In contrast, when elements are referred to as being "directly
connected" or "directly coupled," there are no intervening elements
present.
[0017] The subject matter may be embodied as devices, systems,
methods, and/or computer program products. Accordingly, some or all
of the subject matter may be embodied in hardware and/or in
software (including firmware, resident software, micro-code, state
machines, gate arrays, etc.) Furthermore, the subject matter may
take the form of a computer program product on a computer-usable or
computer-readable storage medium having computer-usable or
computer-readable program code embodied in the medium for use by or
in connection with an instruction execution system. In the context
of this document, a computer-usable or computer-readable medium may
be any medium that can contain, store, communicate, propagate, or
transport the program for use by or in connection with the
instruction execution system, apparatus, or device.
[0018] The computer-usable or computer-readable medium may be for
example, but not limited to, an electronic, magnetic, optical,
electromagnetic, infrared, or semiconductor system, apparatus,
device, or propagation medium. By way of example, and not
limitation, computer-readable media may comprise computer storage
media and communication media.
[0019] Computer storage media includes volatile and nonvolatile,
removable and non-removable media implemented in any method or
technology for storage of information such as computer-readable
instructions, data structures, program modules, or other data.
Computer storage media includes, but is not limited to, RAM, ROM,
EEPROM, flash memory or other memory technology, CD-ROM, digital
versatile disks (DVD) or other optical storage, magnetic cassettes,
magnetic tape, magnetic disk storage or other magnetic storage
devices, or any other medium that can be used to store the desired
information and may be accessed by an instruction execution system.
Note that the computer-usable or computer-readable medium can be
paper or other suitable medium upon which the program is printed,
as the program can be electronically captured via, for instance,
optical scanning of the paper or other suitable medium, then
compiled, interpreted, of otherwise processed in a suitable manner,
if necessary, and then stored in a computer memory.
[0020] Communication media typically embodies computer-readable
instructions, data structures, program modules or other data in a
modulated data signal such as a carrier wave or other transport
mechanism and includes any information delivery media. The term
"modulated data signal" can be defined as a signal that has one or
more of its characteristics set or changed in such a manner as to
encode information in the signal. By way of example, and not
limitation, communication media includes wired media such as a
wired network or direct-wired connection, and wireless media such
as acoustic, RF, infrared and other wireless media. Combinations of
any of the above-mentioned should also be included within the scope
of computer-readable media.
[0021] When the subject matter is embodied in the general context
of computer-executable instructions, the embodiment may comprise
program modules, executed by one or more systems, computers, or
other devices. Generally, program modules include routines,
programs, objects, components, data structures, and the like, that
perform particular tasks or implement particular abstract data
types. Typically, the functionality of the program modules may be
combined or distributed as desired in various embodiments.
[0022] FIG. 1 is a block diagram illustrating a system 100 for
providing application transparent continuous availability using
synchronous replication across data stores in a failover cluster
according to an illustrative embodiment. System 100 includes a
first site "A" 110 and a second site "B" 160. While only two sites
are illustrated in FIG. 1 any number of sites may be present in
system 100.
[0023] The first site "A" (also referred to as the initial primary
site) 110 includes an application cluster resource group 111. The
application cluster resource group 111 includes an application
resource 115, a data disk 120, a log disk 125, and a storage
replication unit 128 (also referred to herein as "primary storage
replication unit 128). These components are associated with the
underlying application that uses the cluster group. The data
generated by this application are stored in these components. The
first site may be a data center that hosts the application
associated with the application resource 115 or may be a server
(physical or virtual) that is hosting the associated application
within a data center or other location.
[0024] Initial primary site 110 further includes a plurality of
nodes 130 and 135. Two nodes are illustrated in FIG. 1 for the
purposes of simplicity only. It should be noted that any number of
nodes may be present on initial primary site 110. Each of the nodes
is associated with an application cluster resource group 111 and
more specifically an application resource 115. Each node 130 and
135 of initial primary site 110 can host a different application
resource. However, in some embodiments the same application cluster
resource group 111 can be hosted on both nodes 130 and 135. In
other embodiments node 135 acts a failover node for the application
cluster resource group 111 on initial primary site 110. When a
failure occurs to an application, the application may failover to
node 135. This can occur in situations where the failure is node
related as opposed to site related. initial primary site 110
further includes a replication service 140 and a physical data
store 150.
[0025] The second or secondary site "B" (also referred to as the
initial replication site or the initial secondary site) 160
includes a replication cluster resource group 161. The replication
cluster resource group also includes a second data disk 170, a
second log disk 175 and a second storage replication unit 178.
Initial secondary site 160 also includes a plurality of nodes 180
and 185. As discussed above with respect to initial primary site
110 only two nodes are illustrated in FIG. 1 for purposes of
simplicity only. Again, any number of nodes may be associated with
initial secondary site 160. Further, initial secondary site 160
includes a second replication service 190 and a second physical
data store 195. The components of the second site "B" 160 are
functionally similar to those of the first site "A" (the initial
primary site) 110, and will not necessarily be described
separately.
[0026] The present disclosure allows for when the initial primary
site 110 that is hosting an application goes down due to a disaster
or otherwise fails, and the application resource group (that also
contains replicated disks for the application) moves to the initial
secondary site 160 that holds a synchronous target replica, that
the cluster physical disk resource to physical data store mapping
is modified to use the target replica, i.e. the replica on the
initial secondary site 160 in data store 195, before the
application resource comes online on the initial secondary site 160
resulting in an automatic role switch. This process is managed by
the replication services 140 and 190. This process virtualizes the
application dependent cluster physical disk resource from multiple
synchronous copies of data in various sites allowing seamless
failover and failback capabilities. Embodiments allow a cluster
replication resource to maintain cluster wide replication state of
all target replicas that allows it to decide if a target is
eligible to be source of replication in the event of a
disaster.
[0027] A target replica connects to source replica without using a
well-known endpoint. When source replica fails over to a different
node within a primary site, the target replica in the secondary
site discovers the new endpoint to connect to and resumes
replication. The cluster replication resource automatically adjusts
the possible owner nodes of the source and target replica based on
the replication state, replication service availability, storage
connectivity, arrival, or departure of nodes in cluster due to
membership changes. This allows the application resource group to
failover to only those nodes where this is a high chance of the
success due to availability of all required resources.
[0028] For the purposes of this discussion, the following terms
will be used to describe the functions of the components
illustrated in FIG. 1.
[0029] Each of the nodes associated with initial primary site 110
can form a replication group such as application cluster resource
group 111 and replication cluster resource group 161. It should be
noted that any number of nodes may form a replication group. A
replication group is in one embodiment a collection of replica
instances on a system that are collectively depended on by an
application using the data partitions of physical data store 150.
The replication service 140 tracks the inter-device write ordering
dependencies when replicating multiple devices. A replication group
is the unit of replication.
[0030] Cluster resource groups: a collection of cluster resources
that are grouped together in a cluster and are a unit of failover
in a failover cluster. These are illustrated by example
application/replication cluster resource groups 111 and 161
[0031] Cluster Physical Disk Resource (PDR): A cluster resource
that manages physical disk so that it can be accessible by
applications, such as applications associated with application
resource 115. Application typically depend on cluster physical
disks resources so that the data is brought online before it can be
accessed by applications
[0032] Storage Replica Resource: A cluster resource that manages
the replication of all replicas in a replication group. Storage
replication unit resources are represented by elements 128 and 178
in FIG. 1 (primary and secondary, respectively). It should be noted
that in FIG. 1, a "p" represents the features on the current
primary site and an "s" represented features on a currently
secondary site.
[0033] Asymmetric Storage Cluster: An asymmetric storage cluster is
a failover cluster deployment where a data store such as physical
disks are not connected to every node of the cluster. Such
deployments are typical found when the cluster spans multiple
geographical sites where physical storage can be accessed by only
the nodes in a given site.
[0034] In order to effectively implement the structure illustrated
in FIG. 1, the replication groups must first be created. The
replicas that belong to a group are grouped together into a cluster
replication group (e.g. application/replication cluster resource
groups 111 and 161). The cluster physical disk resources are part
of replication group that represent the replicas. The replication
groups depend from a cluster physical disk resource, that in turn
depends on the storage replication cluster resource.
[0035] In one illustrative embodiment, the application resource 115
that consume data from physical disks, depend on the physical disk
cluster resource. This dependency chain ensures that the start of
the resources are done in an order that ensures that dependent
resources are available before application can start consuming the
data on the disks.
[0036] The physical data store 150, in one embodiment, is a
physical disc cluster resource that implements shared storage for
the various nodes (e.g. nodes 130 and 135) of the application
cluster resource group 111. Physical disk cluster resources
typically have a private property that indicates the physical data
store 150 it manages. A physical data store 150 in a cluster is
connected to multiple nodes (e.g. nodes 130 and 135) that allows
the data to be available on multiple nodes when the application and
physical disk cluster resource fail over to other nodes. The
physical disk cluster resource takes persistent reservation on
physical data store 150 so that is accessible on only one of the
nodes of cluster to avoid simultaneous edits to data from multiple
nodes. The physical disk cluster resource also mounts the
volume/file systems when it comes online on a node. Collectively
these are illustrated by block 150.
[0037] The replication service 140 is in one embodiment a
replication cluster resource that is configured to determine if
primary storage (physical disks part of physical data store 150
that are part of source replica) is connected to the node where the
resource group is coming online (e.g. application cluster resource
group 111 coming online on initial primary site 110). If the
storage is not connected, the replication service 140 starts an
automatic role switch process. Role switching is a process where
the source and target of replication are switched, thereby
reversing the direction of data replication. Role switching is
typically performed, for example, as part of disaster recovery when
the current primary site (e.g. initial primary site 110) goes down
or when the current primary needs to be taken down for maintenance.
Automatic role switching reverses the direction of replication
automatically when the replication service detects that the current
primary physical storage or nodes are no longer available thereby
providing continuous availability of data to application, without
the need for an administrator to interact directly with the system
during the failover.
[0038] In one embodiment, the replication service 140 determines if
the physical data store 150 is not currently connected by
implementing an associated process. If the physical data store 150
is determined not to be connected, the replication service 140
begins a process to role switch to one of the secondary replication
groups. The process begins by determining if there are other
replication groups (e.g. replication cluster resource group 161)
that can take over as new primary based on their replication
status. If the replication group is in a sync state with an old
primary, then it is also a candidate to be selected as new primary.
Next, the process determines if the replication group is connected
to the cluster node where the resource group is coming online.
[0039] Once a candidate replication group is selected, a role
switch will be performed. This role switch results in the current
primary replication group, e.g. application cluster resource group
111, will be changed to become the target of replication
(secondary). The selected secondary replication group, e.g.
replication cluster resource group 161, will be changed to become
the source of replication (primary).
[0040] To implement the role switch, the process begins by swapping
the binding of the physical disk cluster resource to the physical
data storage of the primary and secondary physical cluster disk
resources. Next, the process swaps the replication group private
property associated with the replication cluster resource. Next,
the secondary cluster resource group is moved to the newly
designated primary site identified in the paragraph above.
[0041] The process continues by updating the possible owners of the
primary and secondary resource groups to include only those cluster
nodes that are within those sites. Possible owners of a cluster
resource are a set of nodes where the cluster can try to bring the
resource online. A replication group may be hosted only on nodes
where the replication service is available, and the physical data
store is available. Again, nodes may be located on different sites.
Additionally, when a primary replication group has synchronous
partners the primary cluster resource group can also be failed over
to those nodes where current synchronous secondary data store is
available. Again, in FIG. 1 a "p" indicates a primary and an "s"
indicates a secondary.
[0042] Continuing with the process of role switching the
replication service tries to online the secondary replication
groups (e.g. application/replication cluster resource groups 111
and 161). This process includes ensuring that the flow of
replication traffic data 145 from the primary site is able to reach
the replication service 190 of the initial secondary site 160. Once
the primary replication resource comes online, the system brings
online the log disk and data disk associated with the primary
resource group, e.g. elements 125 and 120, respectively.
[0043] The replication service 140 maintains a cluster wide view of
replication status of all replication groups within a cluster. The
replications status indicates which replication group(s) are in
"sync" status and which are not in "sync" status. If a synchronous
secondary replication group loses replication connection to its
primary or if there is a failure replicating certain data to a
secondary, the replication status of the secondary replication
group is changed to NOT IN SYNC before the primary deviates or
allows any new input/output to proceed at the primary replication
group to the secondary replication group.
[0044] A replication group can failover within a cluster. When that
happens, the replication to secondary should resume after failover
completes. In the absence of a well-known replication service
endpoint (e.g. a cluster network name resource is an example of a
well-known endpoint), a failover means a change in replication
service endpoint as the node hosting the replication service
changes for the source or target of replication. The older node
that a secondary node was connected to is no longer valid. The
secondary should rediscover the source replication service endpoint
and resume replication. In one illustrative implementation when the
primary cluster resource group fails over to a different node,
during the online sequence it restarts the secondary replication
group. The secondary then during its online sequence queries the
cluster service to determine the owner node of the primary resource
group and uses that node name as primary replication service
endpoint. The primary also sends a cluster resource notification to
secondary replication groups to indicate the new replication
endpoint.
[0045] FIG. 2 is a flow diagram illustrating a process for
selecting a secondary replication group and automatically
implementing a role switch according to one illustrative
embodiment. As the replication cluster resource is at the bottom of
a resource dependency chain, the process begins by determining if
the primary storage 120 is connected to the node. This is
illustrated at step 210. This determination can be made by sending
a command to the primary storage 120 and awaiting a return from the
primary storage 120. Alternatively, the process can be looking for
a heartbeat to come from the primary storage 120. Other methods of
determining if the primary storage 120 is connected to the node can
be used as well. During normal operations, this check occurs when
the associated resource group is first coming online. However, this
check can also occur while the resource group is operating. When
the check occurs while the resource group is operating this check
can occur at periodic intervals or can be a continuous check.
[0046] If the primary storage 120 is determined to be online and
connected to the resource group, the operation of the system
continues as normal. This normal operation is illustrated at step
212.
[0047] However, if it is determined that the storage is not
connected the process moves to step 220 to begin the process of
role switching the storage. At step 220, the process determines if
there any replication groups 178 that can take over as primary
storage replication unit 128. Each candidate is added to a
candidate list. This is illustrated at step 225. Steps 220 and 225
are discussed herein together. In order to be a candidate to take
over as the primary storage replication unit 128, the process looks
to or analyzes the replication status of each of the candidate
replication groups. The replication service 140, 190 maintains a
cluster wide view of the replication status of all replication
groups within a cluster. The replications status indicates which
replication groups are in "sync" status and which are not in "sync"
status. If a synchronous primary replication group 178 loses
replication connection to its primary or if there is a failure
replicating certain data to a secondary, the replication status of
the secondary replication groups is changed to "NOT IN SYNC" status
before the primary deviates or allows any new input or outputs to
proceed at the primary replication group. If the replication status
of the candidate replication cluster resource group 161 is
determined to be in sync with the old primary site that candidate
replication group is consider do be a valid candidate replication
group for selection as a new primary replication group. If it is
not in "sync" with the old primary storage replication unit 128,
that candidate is removed from the list of potential
candidates.
[0048] Once a list of potential candidate replication groups has
been determined, the process continues to determine which of the
candidate replication groups is connected to the cluster node where
the resource group is coming online. This is illustrated at step
230. If the replication group is connected to the cluster node then
that replication group remains in the candidate list. At this time,
the system may gather information about that replication group and
the associated connection between the replication group and the
node or resource group. This information can include features such
as the size or capacity of the replication group, the location of
the replication group, connection speed, quality of the connection,
etc. This information is gathered at step 240.
[0049] If the replication group is not connected to the node the
process moves to step 235. At step 235, the system can remove the
replication group from the candidate list of resource groups.
However, in some embodiments the system can try to have a
connection generated for the resource group to connect to the node.
This is illustrated by optional step 233. Step 233 would typically
occur before step 235. If a connection can be created, the system
causes a command to be generated that will cause the resource group
to connect to the node. If the connection is successful, the
analysis moves to step 240. Otherwise, the replication group is
removed from the list of candidate replication groups.
[0050] Once a final list of replication groups is generated, the
process proceeds to select one of the candidate replication groups
as the new primary replication group. This is illustrated at step
250. The process may select the new primary storage replication
unit based upon the gathered characteristics for each of the
candidates in the list of candidate replication groups. In some
approaches, the selection is based on an existing set of rules for
the resource group. For example, the location of the replication
group may be constrained based on location. This can occur for
certain applications where the data cannot leave a particular
country. In this example, the candidate replication groups not
meeting this location requirement are now removed or not considered
further. The system can look at performance or other quality
characteristics in choosing which candidate replication group to
select. The system may select the best performing replication group
from the candidates. However, the system may simply select a random
candidate from the list of candidates.
[0051] Once the candidate is selected from the list, the process at
step 250 continues by swapping or changing the roles of the two
replication groups. This is illustrated by step 260. At this step,
the current primary storage replication unit 128 is changed to
become a target of the replication. That is, the old primary
storage replication unit 128 is now a secondary replication group.
This is illustrated at step 262. Also, the selected replication
group will be changed to become the new primary replication group.
This is illustrated at step 264. The processes of steps 262 and 264
are essentially the same except for the changes made to the
corresponding replication group.
[0052] The process that occurs on each of the replication groups to
initiate the role switch is described with respect to FIG. 3. The
process of FIG. 3 corresponds closely to step 260 and included
steps 262 and 264 of FIG. 2. The first step of the process is to
swap the physical disk cluster resource to physical data storage
binding of the primary physical disk cluster resources and the
secondary physical cluster disk resources. This is illustrated at
step 310. At this step, the characteristics of the primary physical
disk cluster resources are changed to mimic the secondary physical
disk characteristics and the secondary physical disk's
characteristics are changed to reflect those characteristics of the
primary physical disk cluster resource.
[0053] Next, the replication group private property of each of the
replication cluster resources is swapped. This is illustrated at
step 320. As discussed above every physical disk cluster resource
has a private property that indicates the physical data store it
manages. The physical data store in a cluster is connected to
multiple nodes (e.g. nodes 130, 135 and nodes 180, 185) and allows
the data to be available on multiple nodes so that the application
and the physical disk cluster resource can fail over to other
nodes. The physical disk cluster resource takes persistent
reservation on physical storage so that is accessible on only one
of the nodes of cluster to avoid simultaneously edits to data from
multiple nodes. Thus, the private property of the cluster resource
is changed to accept the edits from this node of the cluster.
[0054] Following the swapping of the private properties and the
binding of the resources the secondary cluster resource group is
moved to the primary site. This is illustrated at step 330. Next,
the possible owners of the primary and secondary resource groups
are updated. This is illustrated at step 340. The possible owners
are updated to include only those cluster nodes that are within
those sites. Possible owners of a cluster resource are a set of
nodes where the cluster can try to bring the resource online. A
replication group can be hosted only on nodes where the replication
service 140, 190 is available and the physical data store is
available. Additionally, when a primary storage replication unit
128 has synchronous partners, the primary cluster resource group
(e.g. application cluster resource group 111) can also be failed
over to those nodes where current synchronous secondary data store
is available.
[0055] Next, the secondary replication group(s) attempt to come
online. This is illustrated at step 350. At this point, the
secondary may have to discover the replication endpoint. If the
secondary resource group knows the replication endpoint, such as a
cluster network name resource, the resource group is able to come
online quickly or with minimal delay. However, during a failover a
change in the replication service endpoint can occur as the node
that was hosting the replication service 140, 190 changes for the
source and/or target of the replication. This occurs as the older
node to which one of the secondary resource groups may no longer be
valid. In some cases, the new endpoint is known. However, if the
replication endpoint is unknown then the secondary resource group
needs to discover the replication end point. This is illustrated at
optional step 360. In this case, the primary replication group 178
rediscovers the source replication endpoint and resumes the
replication process. The primary replication group 178 then as part
of the online sequence queries the cluster service to determine the
owner node of the primary resource group and uses that node's name
as primary replication service endpoint. At this time, the primary
storage replication unit 128 can also send a cluster resource
notification to the secondary replication groups to indicate the
new replication endpoint. Once the endpoint has been discovered and
the secondary storage replication unit 178 has come online, the
failover process is completed. Operation then returns to normal for
the application.
[0056] FIG. 4 illustrates a component diagram of a computing device
according to one embodiment. The computing device 400 can be
utilized to implement one or more computing devices, computer
processes, or software modules described herein. In one example,
the computing device 400 can be utilized to process calculations,
execute instructions, receive, and transmit digital signals. In
another example, the computing device 400 can be utilized to
process calculations, execute instructions, receive and transmit
digital signals, receive and transmit search queries, and
hypertext, compile computer code, as required by the system of the
present embodiments. Further, computing device 400 can be a
distributed computing device where components of computing device
400 are located on different computing devices that are connected
to each other through network or other forms of connections.
Additionally, computing device 400 can be a cloud based computing
device.
[0057] The computing device 400 can be any general or special
purpose computer now known or to become known capable of performing
the steps and/or performing the functions described herein, in
software, hardware, firmware, or a combination thereof.
[0058] In its most basic configuration, computing device 400
typically includes at least one central processing unit (CPU) or
processor 402 and memory 404. Depending on the exact configuration
and type of computing device, memory 404 may be volatile (such as
RAM), non-volatile (such as ROM, flash memory, etc.), or some
combination of the two. Additionally, computing device 400 may also
have additional features/functionality. For example, computing
device 400 may include multiple CPU's. The described methods may be
executed in any manner by any processing unit in computing device
400. For example, the described process may be executed by both
multiple CPU's in parallel.
[0059] Computing device 400 may also include additional storage
(removable and/or non-removable) including, but not limited to,
magnetic or optical disks or tape. Such additional storage is
illustrated in FIG. 4 by storage 406. Computer storage media
includes volatile and nonvolatile, removable and non-removable
media implemented in any method or technology for storage of
information such as computer readable instructions, data
structures, program modules or other data. Memory 404 and storage
406 are all examples of computer storage media. Computer storage
media includes, but is not limited to, RAM, ROM, EEPROM, flash
memory or other memory technology, CD-ROM, digital versatile disks
(DVD) or other optical storage, magnetic cassettes, magnetic tape,
magnetic disk storage or other magnetic storage devices, or any
other medium that can be used to store the desired information and
that can be accessed by computing device 400. Any such computer
storage media may be part of computing device 400.
[0060] Computing device 400 may also contain communications
device(s) 412 that allow the device to communicate with other
devices. Communications device(s) 412 is an example of
communication media. Communication media typically embodies
computer readable instructions, data structures, program modules,
or other data in a modulated data signal such as a carrier wave or
other transport mechanism and includes any information delivery
media. The term "modulated data signal" means a signal that has one
or more of its characteristics set or changed in such a manner as
to encode information in the signal. By way of example, and not
limitation, communication media includes wired media such as a
wired network or direct-wired connection, and wireless media such
as acoustic, RF, infrared and other wireless media. The term
computer-readable media as used herein includes both computer
storage media and communication media. The described methods may be
encoded in any computer-readable media in any form, such as data,
computer-executable instructions, and the like.
[0061] Computing device 400 may also have input device(s) 410 such
as keyboard, mouse, pen, voice input device, touch input device,
etc. Output device(s) 408 such as a display, speakers, printer,
etc. may also be included. All these devices are well known in the
art and need not be discussed at length.
[0062] Those skilled in the art will realize that storage devices
utilized to store program instructions can be distributed across a
network. For example, a remote computer may store an example of the
process described as software. A local or terminal computer may
access the remote computer and download a part or all of the
software to run the program. Alternatively, the local computer may
download pieces of the software as needed, or distributively
process by executing some software instructions at the local
terminal and some at the remote computer (or computer network).
Those skilled in the art will also realize that by utilizing
conventional techniques known to those skilled in the art that all,
or a portion of the software instructions may be carried out by a
dedicated circuit, such as a DSP, programmable logic array, or the
like.
* * * * *